Open zhimin-z opened 1 year ago
Just to make sure if I understand correctly, do you mean the relationship between topics? If so, then you can use cosine similarity between the topic embeddings to see how they relate to one another. You can find something similar in .visualize_heatmap()
.
We know that for any single post, it is possible to calculate the probability distribution, but can we do the same thing on topics? That is, how to calculate the topic correlation after modeling? @MaartenGr